The vertical distribution of clouds has a large impact on the radiative heating and cooling rates of the atmosphere and the surface. Assumptions regarding the vertical cloud overlap in a grid column are required in climate models for the radiative transfer calculations. These various assumptions can lead to large differences in subsequent radiative heating rates of the atmosphere and the surface. Changing the longwave cloud overlap from maximum to maximum-random in the Rossby Centre coupled Atmospheric Ocean regional climate model, RCAO, increases the effective cloudiness by 10%, giving rise to a difference of 1.5°C in two-meter temperatures over northern Europe in wintertime. This also has a large impact of the Baltic Sea ice extent and the length of the ice season.

The cloud overlap assumption can be validated comparing the model output with ground based cloud profiling radar data for particular locations and limited time periods. The RCAO simulated cloud vertical structure and different assumed cloud overlap assumptions are validated using cloud radar data from the EU Framework 5 project CLIWA-NET. Preliminary results show that the observed true overlap can be larger than random overlap, due to the limited horizontal and vertical resolution of the model and the 3D effects of real clouds. The sensitivity of the results to different horizontal and vertical resolutions of RCAO will be presented and the impact on the temperature and the longwave and shortwave surface fluxes will be discussed.